In Nkandanzovu, a recently settled Tonga-speaking region in southern Zambia, a combination of social and environmental pressures contributes to a scarcity of arable land. Informants indicate that they feel pressure to extend their maize fields in order to compensate for a contraction of government fertilizer subsidies. Additionally, family sizes have grown and households have multiplied since the region was first settled, which places further stress on land availability. Finally, over the last two and a half decades of settlement, land has lost much of its fertility under continuous cultivation. Though Nkandanzovu is not a densely populated region, with approximately 172 households spread over a large geographic area, the land necessary to support the low-yield extensive farming practised by local residents is simply not available. As a result, fathers are finding it difficult to allocate land to their sons after they marry; headmen claim they no longer have land to give to needy community members; families are dividing as an increasing number of people leave the region in search of new land; and a growing number of people suggest that a lack of land contributes to their food insecurity. The days of free access to customary land in Nkandanzovu are drawing to a close. In place of free land access, a system of market-based land allocation has emerged. Yet, this market is not fluid and impersonal, but rather is socially ‘jagged’. In this article I argue that local systems of land signification blend with the ‘illegality’ of land markets in Zambia's customary areas to produce a clandestine land market in which women and other marginalized social categories are barred from participation.

Jen Mubita's experience with the emerging land market is emblematic of its jagged nature. Jen is a middle-aged woman who makes a living farming maize, growing garden vegetables, and selling buns in the local market of Nkandanzovo. Divorced by her husband in 2006, Jen, along with her six children, now struggles to produce enough maize to meet the household's food requirements on what she calls her ‘shrinking’ land. Like most people in Nkandanzovu, Jen acquired the land that she currently farms through a combination of her membership within an established homestead and the mechanisms of customary authority. Prior to her divorce, Jen acquired two fields, which she continues to farm today. One of her fields, a small plot of land located near a stream, was acquired from the local headman, while the other, larger, field was given to her by her maternal uncle. While she was married, Jen also worked with her husband and co-wives on the ‘household’ maize fields. During that time, Jen's two fields served as sources of income and occasionally food for her family, while the household's fields provided the bulk of the homestead's maize requirements.

Following the divorce, Jen's ex-husband left the region of Nkandanzovu and returned to his natal home near Lake Kariba. Before he left, he sold the household maize fields to another local farmer, thereby ensuring that Jen and her children could not claim the field and farm it in his absence. He also took with him the majority of the oxen the family once shared to plough the fields. Without draught power, Jen now struggles to plant her fields during the onset of the rains, which seriously reduces the productivity of the maize that she plants. While overall production has decreased on her existing maize fields, her fields are also shrinking as two of her male children approach the age of marriage and are now demanding that she provides them with land to farm independently. This social shrinking of a physically bounded space suggests that the terms by which land is defined as a property are wrapped up in what Carney and Watts call the multivalency of household property rights. In this view, households appear as not merely the loci of competing rights, interests and obligations, but also constitute a complex matrix of nested and overlapping claims to resources, in which structures of property and domination are embedded (1990: 218). Exploring Jen's relationship to the land under her control illuminates this multivalent household structure, and the multiple scales of power relations upon which it rests.

During one of several interviews I conducted with Jen, I suggested that the dual land pressures she currently faced, and the food insecurity it produced, could be remedied by selling the smaller of her two fields in order to generate income to purchase additional oxen and to invest in her bun-making business. As I saw it, with the addition of more oxen, Jen could plant her remaining field early in the rainy season, and thus expand maize production, while also freeing up more time to concentrate on her bun-making and gardening businesses. With any additional income generated by this production shift she could then attempt to purchase new fields for her sons. Her response to my suggestion is illuminating.

Jen explained that buying and selling land is not an option for women, even for divorcees. Her reasoning rests upon two interrelated aspects of property relations and gender in Nkandanzovu. First, the land she controls is not actually hers to sell, though she is emphatic that the land is her possession. Instead, ‘her’ land is more appropriately conceptualized as the future property of her male children. As such, they exercise considerable control over her land, and would never allow her to sell something they consider to be theirs. As another informant suggests, ‘when it comes to land sons have a lot of power’. Second, despite calls by development experts to expand private property rights in customary areas of rural Zambia (World Bank 2004; Roth 1994), buying and selling land in customary areas remains illegal under customary law. Accordingly, the market for land in Nkandanzovo must operate in a clandestine manner, hidden from the gaze of the customary authorities, who reserve the right to expel anyone found to have bought or sold land in their areas. As Jen points out, it is the clandestine nature of this market that prevents her and other women from participation. Ensuring that money spent on land translates into permanent usufruct rights to the land requires mobilizing significant social power to hide the transaction and protect it from any punitive repercussions from the customary authorities. The gendered contradictions of this clandestine market are striking. Although Jen's claims to household land were terminated after her divorce and her ex-husband's subsequent sale of the land, the social power to achieve this level of land alienation is not available to her. Jen feels that she lacks the social power to exercise exclusive control over land she ‘owns’ in order to sell it, and also to extend control over additional land through a purchase. Jen's story is not an isolated one.

Through a series of interviews with men and women over the nature of land and the emergence of a market for land in Nkandanzovu, as well as participant observation over the course of a year, the exclusionary nature of this emerging market became apparent. Yet, as I will argue here, this is not a case of a ‘traditional’ structure of patriarchy simply being extended to emerging markets and relations of property. Instead, the shape and contours of this market are deeply embedded within the multivalent nature of women's control of land and the fractured structure of land governance, in which the national government and customary authorities compete and collaborate over land administration and tenure.

The immediate concern of this article is why the market for land that is developing in Nkandanzovu is so resistant to the impersonal forces that underpin many capitalist market systems. More specifically, why is it that men are so much more capable of exercising the social power to participate in this clandestine market than women? Superficially, it would appear that participation in the illegal and clandestine market for land in Nkandanzovu requires a certain type of social power that is denied to women, what Davison (1988) refers to as ‘gender-structuring.’

Though not a new phenomenon, international concern with land titling in rural Africa gained momentum in the wake of ‘the crisis in African agriculture’, when pictures of starving African children gripped the imagination of the West and demanded a response. Beginning in 1982, the World Bank and other lending institutions dramatically increased funding for land-related projects and began to attach land titling components to their standardized package of conditionalities for Third World debt restructuring (Bassett Reference Bassett, Bassett and Crummey1993). In Africa, development economists argued that by freeing African farmers from the yoke of ‘customary’ land allocation systems and instituting a system of private land tenure, a radical shift in farmers’ relation to the land would take place, thereby creating the conditions for ‘broad-based’ economic growth in rural areas (World Bank 2004; Roth 1994). Although Bank publications (for example, World Bank 2003) more recently began to signal a willingness to countenance a more gradual evolution in land tenure, this earlier economic and ideological pressure is evident in the 1995 Zambia Land Act's emphasis on mechanisms for transferring customary land to individually titled state land. However, the legal changes have been largely ignored and most land in Zambia continues to be managed through customary law under the administration of customary authorities, much as it was under colonial rule (GRZ Reference GRZ1995). The legal pluralism that defines Zambia's fractured system of land governance becomes blurred through the practice of creating and maintaining the clandestine market for land. Thus, the emerging market for land in Nkandanzovu provokes a more profound question about modern nation building in Zambia: what does Jen's exclusion from the market for land reveal about the nature of marginalization in an era in which notions of development are often tied up with an expansion of private property rights, while systems of land administration continue to rest on the historically situated system of legal pluralism?

I argue that within the interstices of this fractured governance regime, land distribution, and more specifically the market for land, becomes subsumed by the multivalent local politics of signification that define land as particular types of property. This politics of signification occurs at multiple scales. Competition between national, international and customary bureaucracies over the terms by which development is enacted, and the ‘appropriate’ form of land tenure in Zambia, articulate with the local politics of land control to produce conditions for the entrenchment and exacerbation of local power differentials and wealth inequality in Nkandanzovu, particularly the further erosion of female control of land.

The analysis presented in this article is the result of over a year of ethnographic study in Nkandanzovu beginning in 2006 and continuing through 2008, as well as archival research conducted at the Zambian National Archives and the University of Zambia. The ethnographic data were collected through a variety of channels: participant observations while living as a member of two homesteads; 105 semi-structured interviews with local residents, which were carried out in English or with the aid of one of three local research assistants; and a structured survey conducted with 72 households. However, given the illegal nature of the land market in Nkandanzovu, many people were reticent about divulging information concerning land markets, particularly as part of a survey, for fear that the information would be shared with the chief. This justifiable fear makes it impossible to provide quantitative data on land sales in Nkandanzovu. As a result, it is people's personal narratives, which emerged over time, coupled with events that I witnessed and explored during the course of my fieldwork, that provide the data for the analysis that follows.

The first section of this article traces the political and economic logic that underpinned the creation of the fractured system of land governance The following section then turns to an analysis of how this fractured system was constructed and deployed in the name of development, with particular attention paid to the historical erosion of women's rights to land in Tonga-speaking regions. The article then proceeds to an analysis of land taxonomies in Nkandanzovu and their relationship to the norms that govern land allocation. Far from a stable set of norms and behaviours, the local practices of land distribution are shown to rest on a system of signification, in which the local social power to define land in a particular way influences the ways in which it is distributed. The next section explores the contemporary laws governing land administration in Zambia, particularly the inclusion of mechanisms to convert customary land to state land in the most recent Zambian Land Act. This policy change articulates with the historically situated structures of customary rule and the local politics of signification that regulate land distribution to produce a clandestine land market governed by the rules of formalized illegality. The mechanisms of formalized illegality are shown to allow some men to profit from and consolidate land under their control, while denying women the power to participate. The article concludes by examining the implications of this system in an era of increased incidences of divorce and widowing brought about by the spread of HIV/AIDS.

THE HISTORICAL EMERGENCE OF A FRACTURED SYSTEM OF LAND GOVERNANCE

In 1928 the British Colonial Authority, through the Council-in-Order, divided the territory of Northern Rhodesia between crown land and native reserves. Crown land was for the occupation of white settlers and was to be administered under British and statutory law. Within these regions, which included a vast swathe of the most productive land in the Tonga Plateau region of Southern Province, the colonial government granted freehold or leasehold title to white settlers. This land designation was to serve as an incentive to attract European settlers to the region. Conversely, native reserves were set aside for African people, to enable them to practise, in perpetuity, their ‘traditional’ way of life. Land in these regions was vested with the ‘chief’ for the benefit of the ‘tribe’. The ideological foundation of this system of land governance was a central component of British indirect rule, originally explicated in Lugard's Dual Mandate for Tropical Africa (1923).

Indirect rule emerged in British colonial Africa out of political and economic necessity, though moral justification was also found in ‘preserving’ African cultural integrity from the adverse affects of external rule. In 1924, when the British Colonial Office took control of the territory of Northern Rhodesia from the British South Africa Company, the colonial government was confronted by the dual and contradictory demands of extracting a profit from its colonial holdings, while at the same time attempting to minimize the socially disruptive effects of their interventions (Berry Reference Berry1993: 29). In an effort to ameliorate these contradictions, colonial authorities sought to integrate existing local authority and political systems into the colonial system of administration and rule. To facilitate the absorption of indigenous political systems, colonial officials sought to understand how African political systems functioned prior to colonial interventions.

Anthropologists and sociologists, such as Audrey Richards, Max Gluckman, Meyer Fortes, A. R. Radcliffe-Brown and E. E. Evans-Pritchard, were employed by colonial governments to study and define the terms of traditional indigenous political organization. From their work, an anthropological perspective on ‘primitive societies’, known as structural functionalism, emerged. This perspective challenged many of the then dominant evolutionary perspectives on ‘modern’ versus ‘primitive’ societies, which saw Africans as below Europeans on a scale of social and cultural evolution. Through intensive fieldwork, structural functionalists sought to define and explicate the logic of African social systems and indigenous morality, thereby valorizing African societies for Western audiences (Falk Moore 1994: 23). However, many of these functionalists presupposed a closed system, in which African societies had existed as stable, unchanged, integrated and coherent social wholes before the disruption of colonial rule. This interpretation of African societies, as Falk Moore (1994: 24) argues, ‘fit[s] neatly with the political conception of the “tribe” used by colonial administration to divide up the population it governed into intelligible units’. The creation of the “tribe” allowed colonial officials to transform the variegated social and political landscapes of their Africa colonies into territories that were ‘legible’ to their rule and extractive interventions (Scott Reference Scott1998).

From this conceptualization of the tribe, policies were developed that granted authoritative power over the administration of tribal communities to local elites or chiefs. However, among many Tonga people in southern Zambia chiefs were not an historical fact, but rather a colonial invention used to create indigenous allies for ruling and administering the territory (Berry Reference Berry1993: 28–30). As anthropologist Elizabeth Colson comments, the Tonga people of the Gwembe Valley ‘regarded the chieftaincy as a purely arbitrary creation of European officials’ (1971: 29). Similarly, in their auxiliary work for the colonial Native Land Tenure Committee, Gluckman et al. note that ‘we doubt if the Tonga ever had chiefs as these are usually defined’ (1948: 58). Furthermore, as many Africanist scholars have shown, African societies were not comprised of traditional units of social organization, such as tribes (Berry Reference Berry1993; Hobsbawm and Ranger 1983). In attempting to (re)create a definition of traditional practice through indirect rule, colonial officials, with the aid of anthropological research, codified a set of norms based on synchronic interpretations of custom, funnelled through a European understanding of political organization. Thus, far from stabilizing the terms of African customs, as some have argued (Ranger Reference Hobsbawm and Ranger1983), the codification of customary law provoked a series of debates over the very terms of traditional behaviour (Berry Reference Berry1993: 24; Ferguson 1997). Upon this contested terrain, policies aimed at developing rural Zambia were inserted.

DEVELOPMENT, GENDER AND THE DEFINITION OF CUSTOM IN SOUTHERN ZAMBIA

Following the global economic depression of the 1930s a radical rethinking of the role of the state in promoting and maintaining economic stability was undertaken in Britain, with important consequences for the administration of its colonies. As Berry states, ‘by the 1940s, Keynesian economics had so far penetrated the Colonial Office that … there was even talk of a Marshall Plan for the colonies’ (1993: 47). In 1940 the Colonial Office promulgated the Colonial Development and Welfare Act, which ‘increased spending on development schemes and conservations works, and mapped out ambitious blueprints for improved African farming’ (Berry Reference Berry1993: 47). Having already created a fractured system of land governance in Northern Rhodesia, the Colonial Office sought to utilize the structures of customary law in an effort to promote Improved African Farming Schemes in Tonga areas. To this end, anthropologist Max Gluckman, along with several agronomists and colonial officials, undertook a study entitled Land Holdings and Land Usage among the Plateau Tonga of Mazabuka District in 1945. This study sought, on the one hand, to define the terms of Tonga customary law, and, on the other, to suggest modes of development intervention that fit with this interpretation of custom. In this report, Gluckman et al. describe customary Tonga land holdings as follows:

They go on, taking a decidedly gendered view of customary land administration, to note that:

From this perspective it appears that (1) land holding is customarily the domain of men, though, in practice, women controlled a quarter of total cultivated land; (2) Tonga people are essentially egalitarian, organized around the principles of community support; and (3) the practice of shifting cultivation and grazing needs dictate what constitutes sufficient land. Given this understanding of Tonga custom, the authors then go on to refute claims being made by Tonga people at the time, who stated that soil erosion in the reserves, which was a major preoccupation of the colonial regime, was caused by the displacement of Tonga people from the most fertile land in their territory onto Native Reserves. Instead, the authors suggest ‘primitive’ farming methods (that is, shifting cultivation), coupled with low expectations for yields and quality of life, are the cause of Tonga people's land troubles, what they call the ‘Tonga problem’. Having thus interpreted the customary law of Tonga people, they locate both the cause and solution to land issues within this conceptualization of custom. As such, they recommend using ‘tribal communities rather than families as the unit for agricultural development’ (ibid.: 6), which they later call social-agricultural units, to facilitate radical changes in Tonga people's approach to land and agriculture without requiring a ‘serious break with traditional land tenure’ (ibid.: 8).

This gendered conceptualization of land tenure among Tonga people as an instrument of agricultural development and rural governance contrasts markedly with the analysis of Elizabeth Colson, and her work among both Gwembe and Plateau Tonga (Colson Reference Colson1958; Colson 1960). The Gwembe Valley, which was the original home of many of the migrants in the Nkandanzovu area, once supported numerous small communities on alluvial gardens on the edge of the Zambezi River. These alluvial fields could be farmed indefinitely, as their fertility was restored during annual flooding (Colson Reference Colson1971: 72; Cliggett 2005: 64). Matrilineal lineages were strongly defined among Gwembe Tonga, ‘presumably due to the fact that the alluvial fields which could be cultivated indefinitely were highly valued and subject to inheritance. As a result, matrilineal kinsmen, who were one another's heirs, shared vital property interests more conducive to lineage cohesion’ (Colson Reference Colson1971: 72). Inheritance of land among the Gwembe Tonga was linked to what Colson calls ‘the cult of the shade of the dead’ (Colson Reference Colson1960: 122–44). Through this system, the death of a lineage member would require the appointment of another member to inherit the shade (muzimu) of the deceased, along with the rights that the deceased person exercised while alive, including rights to land (Colson Reference Colson1971: 72). Generally, men would inherit from men and women from women; thus women and men were both in a position to inherit valuable river land. Children did not expect to inherit land from their fathers, though fathers did give children gifts of land and stock (ibid.: 73). Instead, inheritance of land occurred through matrilineal relationships, which often provoked serious quarrels and even murder between competing kin (ibid.: 74; Colson 1960).

Thus, far from the community-based, egalitarian, and gendered vision of land allocation described by Gluckman et al. (1948), Colson paints a picture of a system in which men and women had access to land, while disputes over land were a common part of daily life.

This is not to say that Gluckman et al. were entirely wrong in their interpretation of land tenure. As Colson demonstrates in her longitudinal study of Gwembe Tonga people, following the construction of Kariba Dam in the late 1950s, and the subsequent displacement of thousands of Tonga people from their alluvial gardens along Zambezi River to the Batoka Plateau, shifting cultivation on large tracts of land replaced the permanence of alluvial fields, thereby changing Gwembe people's relationships between each other and the land. Disputes between kin over land decreased as land became more available and less permanent, and thus less desirable as an inheritance (Colson Reference Colson1971: 74). Instead, disputes over the inheritance of stock, particularly oxen to plough new extensive fields, began to supersede disputes over land. Additionally, women's rights to land were degraded by the importance of clearing the land in the new settlement areas on the Plateau.

Throughout Tonga country, rights to land are often tied to clearing the land of trees and brush. Though nothing prevents women from clearing land, or paying to have it cleared, clearing land is generally considered a man's job. As such, following displacement from the Gwembe Valley, men in the new settlement areas on the Plateau increased their control over land by being responsible for its clearing. At the same time, ownership and control of land is transferred to individuals who clear land, and is not necessarily open to the same claims by kin as permanent alluvial fields that do not require clearing (Cliggett Reference Cliggett2005: 68). As such, land access following displacement came further under male control and became more distanced from multiple claims from kin. Additionally, systems of inheritance have changed to reflect changes in ownership and social organization. With the material basis of clan cohesion, the alluvial fields, destroyed, children have increasingly come to inherit land and other resources from fathers rather than matrilineal kin (Cliggett Reference Cliggett2005: 66). In many ways, as Colson (1971) notes, the system has come to mirror quite closely the system of land allocation practised by other Tonga people on the Plateau (ibid.: 90, see also Colson 1958). Yet this is clearly not the same system described by Gluckman et al. (1948).

Individual ownership and control over land through family relations, rather than community access, is the norm among Tonga people who practise extensive farming. While extensive fields on the Plateau are predominantly controlled by men, women can and do acquire land through inheritance, payment for clearing land, and gifts. This is not an aberration of ‘custom’, but rather is part of the system by which land and resources are distributed among Tonga people. Given the clear evidence that Tonga farmers on the Batoka Plateau tend to control resources, including land, individually and across gender lines, a fact that Gluckman et al. (1948: 8) recognized, why would they insist on a definition of Tonga customary law that seems so far removed from the system they observed?

The gendered and ‘community-based’ understanding of Tonga customary law has its roots firmly planted in efforts to promote a very particular notion of ‘development’ in southern Zambia. As G. Beresford Stooke, an ‘expert on the administration of Africans’, stated in the 1940s:

Clearly the idea of community development has a long and unsavoury history in Zambia. Following Stooke's concern for promoting the community as the unit of development interventions, Gluckman et al. (1948: 6) recommend that the problems of a rising resentment among Tonga people toward colonial rule, particularly following their displacement from the most fertile land of Southern Province to native reserves, and massive soil erosion in the overpopulated reserves, ‘can be overcome, at least partially, by using communities rather than families as the unit for agricultural development’.

For the colonial authorities, ‘communities’ had a very specific meaning. As a unit of agricultural development, communities were to be comprised of ten to fifteen tax-paying males, or ‘heads of household’, under the administration of a headman and eventually a chief (Gluckman et al. 1948). Through this arrangement, taxes could be collected easily, land could be divided equally between men, and development interventions, particularly the Kanchomba system of crop rotation and green manure, could be implemented (Johnson Reference Johnson1956). Within these administrative communities, individuals (that is, males), would have a right to an equal share of land and have full rights to its products. Women, however, would not be granted the same rights, as they were not considered tax-paying heads of households, nor were their rights to land considered customary. Instead, as we have seen, female land holders were seen as obstacles to agricultural development and customary social cohesion (Gluckman et al. 1948: 4)

This gendered system of land allocation was facilitated by the chief and other customary authorities. Additionally, chiefs became the conduits for enforcing ‘improved’ agricultural practices, such as contour ridging and cassava gardening, within their reserves. Any chief or headman who resisted being an ‘agent of improvement’ would be removed and replaced by someone more amenable to colonial directives (Chipungu Reference Chipungu1988: 72). Yet the position of chiefs as ‘agents of improvement’ further undermined their authority, as many Tonga people resented having their farming practices regulated and imposed upon (Colson Reference Colson1971). This historically situated relationship between gender, land governance and development continues to shape the ways in which people in Nkandanzovu access land today.

CUSTOMARY RULE AND LOCAL NORMS ON LAND CONTROL AND ALLOCATION

Despite historical efforts to undermine women's rights to land for the sake of development and nation building, women in Nkandanzovu continue to control land today. However, more so than men, women's rights to and control of land are tied up in a complex matrix of overlapping and concurrent, or ‘bundles’, of property rights (Berry Reference Berry, Downs and Reyna1988; Carney and Watts 1990). In many ways, struggles over control of land rest on the politics of signification that regulates the ways in which land is distributed through a system of land taxonomy. Generally speaking, women's land tends to fall within broader categories of land that leave it open to a plurality of claims, particularly from sons and husbands. Because of the multiple claims that can be levelled against women's land, their individual control of land is always tempered. Furthermore, the existence of multiple claims to women's land tends to prevent women from participating in the emerging clandestine market for land, which must be hidden from customary authorities.

Like men, women gain access to land by being given land by a relative, by being allocated land by customary authorities, and through inheritances. Additionally, divorced or widowed women often obtain land to farm from their married sons, while married women are frequently given land by their husbands in order to generate income for household goods that are considered a woman's responsibility. The primary differences between women's and men's rights to land are tied up in conceptualizations of gender roles in Tonga society. In general, men, as husbands, exercise considerable control over their wives, for whom they have paid a bridewealth (labola), but responsibility for children ultimately rests with their wives. Women derive much of their social position from their role as guardians and providers for their children. The position women hold as guardians of their children and ‘property’ of husbands undermines their ability to exercise exclusive control over land.

Broadly speaking, land allocation and control in Nkandanzovu can be divided into two forms of usufruct rights, temporary and permanent. The distinctions between them rest in large measure on the ways in which social categories – including gender, age, and community insiders versus outsiders – are enacted in daily practice. Permanent usufruct rights to land are granted to individuals in several ways: (1) fathers will often give land to their sons prior to their death in order to remove that land from the possibility of an inheritance dispute; (2) inherited land is often considered permanent and secure, although the process of inheritance is open to debate and struggle; (3) clearing land of trees or paying for land clearing generally conveys individual ownership and permanent control; and (4) permanent rights to land are granted through the mechanisms of the customary authority.

Women tend to be denied permanent control of land for several reasons. First, the process of bridewealth payment, which is paid by the family of the man to the family of the woman, formally separates women from their natal families. Any resources under the woman's control would be transferred, therefore, to her husband and his family. Thus fathers rarely give land to their daughters, because that land would ultimately wind up in the hands of her future husband. Second, although women's rights to inherit land from their husbands were strengthened under new national inheritance laws promulgated in 1989, the land women inherit is thought to be for the benefit of their children, and is not considered an individual asset. This has important implications for women's ability to participate in the emerging land market. For example, one widowed woman's attempt to sell her inherited land to a wealthy farmer was stopped by the uncle of her children, because, as he argued, ‘that land is not for you to sell. That land is for your children.’ Third, clearing land is considered a man's job. This gendered division of labour prevents women from gaining access to fertile chisaka, or forested land, and denies them one of the most powerful mechanisms for obtaining and controlling land in Nkandanzovu. Finally, with few exceptions, customary authorities give land to men, as heads of household, while assuming women will gain rights to land through the ‘conjugal contract’ with their husbands (Whitehead Reference Whitehead, Young, Wolkowitz and McCullagh1981).

Since women are denied access to the mechanisms for controlling land on a permanent basis, land under their control tends to fall within the broad category of temporary usufruct rights. Temporary land in Nkandanzovu is allocated in several ways: (1) wives are often given a small portion of their husband's land upon which to cultivate ‘women's crops’, including groundnuts, sweet potatoes and cow peas. The purpose of these fields is to provide women with the means to generate income for purchasing kitchenware and clothing, as well as to augment their family's diet. (2) New arrivals to the community are often given land to farm temporarily by an established member of the community while they search for permanent land. The temporary gift of land to an outsider is known as chilianyenda, which is roughly translated to mean ‘passing through for a meal’. Thus, chilianyenda is given to help an outsider feed themselves as they establish their own farms. (3) Sons-in-law are occasionally given land temporarily by their fathers-in-law in cases where the son-in-law, for whatever reason, has not received land through his father. It is also important to note that because clearing land of trees conveys ownership, forested chisaka cannot be given temporarily, because to farm it requires clearing the trees.

The mode of allocating temporary usufruct rights to land differs markedly from how permanent rights are granted. Permanent usufruct rights are transferred through locally formalized mechanisms that involve the village committee, while temporary land use is granted within the context of the homestead. The village committee is comprised of the village headman and a cadre of local village elders. The committee is normally convened in order to survey land that is being permanently transferred as a gift to a son, as an inheritance, or, increasingly, through a sale. During the surveying process, the committee becomes familiarized with the land in question and will mark the boundary by making axe marks, or impa, on the trees that form the border. In this way, the transaction becomes formalized and secured, though boundary disputes may arise in the future.

It is important to distinguish village headmen and their committees from the mechanisms of customary rule. Though village headmen are part of the hierarchy of customary rule, their social power is primarily derived from their position as respected members of the local community. It is this relationship to the community, rather than to the broader institution of customary rule, that allows village headmen to participate in illegal land markets. Given the historically adversarial relationship between Tonga people and customary authority, village headmen tend to operate in a quasi-autonomous fashion distanced from chiefly authority. Thus they can be enrolled, and often participate directly, in the emerging market for land without undermining their own authority. Indeed, by supporting the market for land, village headmen stand to enhance their positions within the community by helping powerful members of the community to disguise an illegal land sale from the punitive gaze of the chief. As will be discussed in the following section, disguising an illegal land sale requires reconfiguring the transaction to fit the accepted norms that govern the transfer of permanent usufruct rights – such as a gift, inheritance, or, more commonly, clearing the land of trees.

The temporary transfer of usufruct rights to land, on the other hand, never involves the formal structure of the village committee. Instead, the allocation of temporary rights to land occurs within homesteads. For example, land within a man's field is set aside for cultivation by his wives, without officially marking the boundary separating the two. Instead, the crops themselves mark the boundary, which gives women rights to that land and its produce until the crops are harvested, at which time the land reverts back to male control. Similarly, chilianyenda rights are granted within a cleared field and are maintained until the field is harvested. This prevents an outsider from claiming permanent rights to the land. Thus, temporary land tends to be quite elastic and mobile, as it changes size and location from year to year based on the whims of the field's ‘owner’.

Overall, women tend to obtain only temporary rights to land as part of a conjugal contract with their husbands. Because of the temporary nature of women's access to land, they are denied entry to the formal mechanisms of the village committee, which is a necessary precondition for participation in the local land market. In cases where women are granted land on a more permanent basis, such as through inheritance, this permanence is always mitigated by gendered visions of land ownership, in which women are seen to control land solely for the benefit of their children and not as an individual asset. Thus, while a transfer of land through inheritance or gift is more permanent than land accessed through husbands, women do not exercise individual rights over how to use that land in the same way that men do. Instead, women's decisions over land are often dictated by their male children and other male relatives. One older man, who I knew to be involved in preventing his widowed sister from selling a portion of her land, told me that ‘we (men) must protect the children from women's foolish decisions’. Because of the clandestine nature of the emerging land market, which the chief sees as a direct threat to his authority, coupled with the multiple claims to women's land and the temporary nature of women's land control, women tend to be denied access to the emerging and lucrative market for land as both buyers and sellers. Within the context of these local politics of signification, and the continued predominance of a fractured regime of land governance, policies to promote individual land tenure in rural Zambia have been inserted, to produce a deeply gendered, localized and illegal market for land.

FORMALIZED ILLEGALITY AND THE LOCAL FRICTIONS OF A CLANDESTINE LAND MARKET

In response to calls by development organizations and international lending institutions to expand private property rights in customary land areas, the 1995 Zambian Land Act includes mechanisms for transferring customary to state land (Mudenda Reference Mudenda2006). New institutions, including the Land Development Fund, were created under this legislation to aid in the legal removal of land under customary control to form bounded settlement areas regulated under state law. Through this legislation, pockets of state land have been formed within the boundaries of customary areas. One such settlement area has been created close to Nkandanzovu, in the customary land area of Chief Chikanta. Yet, rather than aiding in the smooth transition from communal to individual land tenure regimes, as the policy envisages, the emergence of this settlement area has intensified long-standing conflicts between the many residents of Nkandanzovu, for whom a land market is still denied, and the chief.

Access to land in the settlement area requires that a significant payment, or ‘tribute’, be given to the chief, who then loses control over the land once it is formally transferred to state land (Brown Reference Brown, Evers, Spierenburg and Wels2005). While residents of Nkandanzovu are frequently threatened by the chief with forced eviction for buying and selling land, this tribute is viewed by the people of Nkandanzovu as tantamount to selling land. Furthermore, many of the beneficiaries of these legal land transfers are outsiders from nearby urban centres, as well as prominent local residents, including the agricultural officer, several headmen, and the veterinary officer. Why, the residents of Nkandanzovu ask, is it legal for the chief to sell land, while we cannot? With growing land scarcities, the creation of the settlement area serves as the primary justification for the emergence of the clandestine market for land in Nkandanzovu. People who participate in the land market often point to the fact that the chief is selling land in the settlement area to justify to themselves and others their participation in the illegal land market. Yet due to the continued illegality of the market for land, this emerging land market must be hidden from the punitive gaze of the chief. To accomplish this, land sales are funnelled through local, ‘accepted’ channels for transferring permanent usufruct rights to land. In particular, the institution of the village committee is mobilized to disguise land sales and to facilitate the illegal market for land. Yet given the gendered politics of significations that regulate land allocations in Nkandanzovu, this system of formalized illegality tends to privilege men, who have the social power to define their land as permanently and exclusively controlled, while excluding women. Thus, while the 1995 Land Act has codified mechanisms to promote market-led development in the customary land areas of rural Zambia, local frictions, including resistance by the customary authorities to ceding control of land and the local gendered politics of land distribution, serve as barriers to a fluid land market. Instead, the market is socially jagged in ways that make it accessible only to those with sufficient social and financial capital to mobilize either the formal process, under chiefly control, of transfer from customary to state land, or the clandestine land market operated through the village committee.

The maintenance and perpetuation of the illegal market for land in Nkandanzovu requires, among other things, that participants in the market feel secure that the transaction permanently transfers usufruct rights to land and that the transaction can be successfully disguised from the punitive gaze of the chief. Consequently, alternative claims to the land must be silenced before a transaction can take place, as these alternative claims may find a receptive ear with the chief, and the mechanisms of permanent land allocation must be enrolled in order to enhance the security of the transaction. To navigate these dual demands, the land in question has to be considered under the exclusive and permanent control of the seller, and the transaction itself must be plausibly disguised within accepted norms of customary land allocation described above. For example, a woman could never sell the land she obtained from her husband, because that land is not considered under her permanent and exclusive control. Though it was given to the woman to farm and meet her individual needs and responsibilities, the land itself, though not the crops, is still considered the property of the husband. Similarly, a woman's attempt to sell inherited land can be challenged by her male children and other relatives, who see that land as the property of her children.

Women face related obstacles when seeking to buy land. Women tend to obtain rights to land in a temporary fashion and are therefore denied access to the formal structures of allocating permanent land that are needed to disguise a land purchase. Moreover, the multiple claims levelled against women's land, particularly by male relatives, make it impossible for women to purchase land as an individual asset. Conversely, men exercise considerably more freedom over land decisions. In one case, a wealthy farmer successfully sold a piece of land to another farmer, though the land in question had been farmed for some years as chilianyenda by a recent migrant to the region. The seller was able to stifle the alternative claim to the land from the recent migrant because, as chilianyenda, the land was never formally recognized to be under the permanent control of the migrant. Additionally, the sale itself was disguised through the mechanisms of the village committee. As one committee member explained to me, what was exchanged was not money for land, but rather money as payment for the previous clearing of land. Having thus paid two million kwacha (roughly US$500) as payment for the clearing of land, a price that far exceeds the price of physically clearing trees, the buyer was awarded permanent rights to the land.

The formalization of these illegal activities through local institutions is often further solidified by writing contracts for land sales and transactions. These contracts describe the land in question and the participants in the transaction, and are signed by witnesses from the village committee. However, the contracts never mention the terms of the transaction itself, including the price and the mechanisms by which the land changed hands. This contract thus provides an additional form of security to an illegal land transaction, by providing material recognition of the transaction itself, while at the same time disguising the terms by which the land changed hands. This is an increasingly common practice throughout Africa. For example, Lavigne Delville notes that farmers in Francophone Africa ‘put transactions on paper not so much to make the content of the transaction comprehensively and unambiguously clear, as to certify that they have indeed taken place’ (2003: 103). The irony of using formal mechanisms to solidify an illicit transaction of land is that the distinctions between formal and informal, as well as state and local, become increasingly blurred in practice. While props of the state, such as land contracts, are produced to secure land transactions, these are generated through local institutions which seek to disguise the transaction from the bureaucratic gaze of customary rulers and the central government. As Benjaminsen and Lund note, this informal formalization ‘produces land tenure systems which are neither regulated by predictable rules and structures nor characterized by sheer anarchy’ (2003: 3). Yet the illegal nature of such transactions, coupled with the systems by which land can be transferred legally, does produce certain predictable outcomes in Nkandanzovu.

CONCLUSION

The most predictable result of the land market in Nkandanzovu is that women are denied access to it, although they control and farm significant portions of the land. It could be said that this protects women (and their dependent children) from loss of their land through distress sales, but equally they are prevented from expanding their land holdings and profiting from what land they control because of the social logic by which the market for land is regulated. This logic rests upon the fractured foundation of land governance regimes in Zambia, including local, customary, and national institutions of land administration, which actively compete over the appropriate modes of allocating resources, the definition of property rights, and the legitimate right to govern. It is within this murky context of governance that questions of development, and particularly ideas about achieving development through the expansion of private property rights, are subsumed. Though encouraged by the opening of settlement areas, the people of Nkandanzovu lack formal support from the national government to create a market for land. At the same time, efforts to create a transparent market for customary land face fierce resistance from the chief who exercises punitive control over land but lacks other forms of local authority. Under these conditions, the market system of land allocation is not free and impersonal, but rather deeply embedded within local power structures that are forced to operate in a clandestine, yet secure manner. To navigate this contradictory terrain, local institutions of land allocation are enrolled and alternative claims to land are stifled. As we have seen, the multivalent nature of women's land control leaves ‘their’ land open to alternative claims from husbands and sons. This makes it exceedingly difficult for women to exercise exclusive rights to the land necessary to complete a sale. Additionally, because of the illegal nature of the land market, enrolling the village committee into a land sale requires that the members of the committee stand to gain either socially or economically from their participation. By and large it is men who control wealth and the means for creating wealth in Nkandanzovu. Thus, while women exercise control over valuable assets, such as land, they are denied alternative means for profiting from them, such as through selling land, or acquiring permanent rights via the market. In this way, land as well as social power and wealth manifest growing inequality between women and men in rural Zambia. This is not the outcome of traditional behaviour stifling the expansion of markets and private property, and thus development. Instead, it is the direct outcome of the fractured nature of the governance regimes through which development is instituted, and the local power differentials that become entrenched within the interstices of these fractures. Women's access to land – whether through market or non-market processes – is in this respect a manifestation of those power differentials.

In Zambia, the average life expectancy for men has dropped from 50.4 years in 1980 to 48.0 years in 2000, while for women the figure has dropped less dramatically from 52.5 years in 1980 to 52.0 years in 2000 (ZDHS 2007: 2). Much of this decline can be attributed to the spread of HIV/AIDS. In Southern Province, HIV prevalence currently stands at 14.5 per cent, with women exhibiting a higher prevalence than men (ZDHS 2007: 237). Today in Nkandanzovu issues of health and death, and divorce and widowing, intersect with tensions over resource allocation and control to produce a deeply troubling situation for women.

Chronic illnesses associated with a depressed immune system can be debilitating. In Nkandanzovu, women who are chronically ill are often divorced by their husbands, as they can no longer maintain the work burden expected of them. As the preceding discussion has shown, divorced women find themselves in an extremely difficult situation vis-à-vis land access and control. Denied access to land markets, divorced women are increasingly unable to acquire land to farm and the resources necessary to generate an income. Lacking a means for acquiring permanent land to farm, divorced women are left dependent on their male relatives, particularly their sons, for their very survival. A similar situation prevails for widows, who may inherit a small tract of land but cannot profit from the sale of this land or expand their holdings over more land through a purchase. Left with little recourse, divorced and widowed women must attempt to eke out a living on small and often shrinking land. Until such time as women can freely participate in local land markets, and can openly challenge the multivalent structures of property and domination imposed on them, women will be trapped within a cycle of under-production and male dependence that is extremely difficult to break.